Illegal dumping has become one of the serious environmental problems. Efficient illegal dumping proven programs must be customized to control the problem. This paper is intended to propose a project in which data mining techniques is employed to extract rules from large amount of spatial data including the location of illegal dumping sites and relative GIS maps, and then the extracted rules are applied to predict the allocation of illicit dumping probability in the research area. Besides, multi-temporal remote sensing images are analyzed to extract the spectral pattern of dumping for detecting undiscovered dumping sites. The project has been conducted in Taoyuan, Taiwan. The effort indicates that the integration of spatial data mining and remote sensing is assistant for monitoring illegal dumping and helpful for protecting our living environment against pollutions.